Earth Observation refers to the collection and analysis of data about the Earth’s surface, atmosphere, and oceans using remote sensing instruments aboard satellites, aircraft, and ground-based systems. Data types include optical imagery, synthetic aperture radar, and multispectral and hyperspectral measurements. Applications span climate monitoring, land use mapping, disaster assessment, and infrastructure analysis. In AI research, earth observation drives work in large-scale geospatial modeling, change detection, and foundation models for remote sensing data.
Satellite imagery is vast, high-resolution, and rich with information, but finding specific objects within it using natural language has remained a stubborn challenge. Open-SAT, developed by researchers at NEC Laboratories America and North South University, tackles this problem without retraining any models.